Metallica - Death Magnetic

[SouthSIDE Glen]

That will be their artistic choice but at least the rest of us won't be forced to do the same just to keep bands and listeners happy in their desire to have music that sits nicely alongside other music.

Sure we will, that's the whole point. The only difference between an RMS of -5 with a crest factor of 5 and an RMS of -14 with a crest factor of 5 is the amount of amplification needed to bring it up to a certain playback SPL level. If we have a peak RMS of -14 with a crest factor of 14, our stuff will have exactly the same problem "competing" for volume, and the smashed mixes will still sound like shit. Nothing will be solved.

I guess we'll have to agree to disagree about this

Yeah, because frankly nothing about the idea of standardization around arbitrary volume levels makes any sense to me on any level. Especially when compared to the idea of actually retaining already proven engineering technique to achieve the exact same desired results.

There's nothing broken with the system. Only with the users of that system. The fix needs to be implemented on the user side, not the system side.

It doesn't much matter anyway, iqi; neither one of us will get our wish. As long as the engineers bitch to each other but only smile at their paymasters, nothing's going to change. Not until the next generation of paymasters comes along, anyway.

Then again, by then there will be no engineers left anyway. All composing, track creation, mixing and mastering will be automated by software. Just push about 5 preset buttons for a few parameters, just to make the user feel important even though they will have already stepped into irrelevancy at that point, and sit back, and everything will come out with everything including the shrinkwrap.

G.

 

[Superhuman]

Gwaaaaarrrrghhhhhhhhh!!! That's the sound of me taking my size 13 foot out of my mouth. I take back all that I said above about this album sounding ok. I just listened to one of the better tracks 'The Day That Never Comes' in the quiet of my parked car and there is major distortion and mud at the end of that track (I had only ever listend to it in a noisy environment before). At first I thought there was something wrong with my cd player, but yea, its there allright - how the hell could the label allow that to be released?? I'd say Metallica are pissed, it couldn't have sounded like that in the control room... could it??
Here's a link to where the engineer who did the mastering complains about what he had to work with. It also show the difference between the Guitar Hero version and the cd version: http://www.nme.com/news/metallica/39816

 

[iqi616]

Sure we will, that's the whole point. The only difference between an RMS of -5 with a crest factor of 5 and an RMS of -14 with a crest factor of 5 is the amount of amplification needed to bring it up to a certain playback SPL level. If we have a peak RMS of -14 with a crest factor of 14, our stuff will have exactly the same problem "competing" for volume, and the smashed mixes will still sound like shit. Nothing will be solved.

Now you're confusing me. Our stuff with RMS of -14 and crest factor of 14 will sound approximately as loud as stuff with RMS of -14 and crest factor of 5. I say approximately because the tonality of the piece affects the perceived loudness but unless the tonality is completely out of wack it should be within a dB or two. Note that the mastering engineer may level a track so that it is quieter than the limit so that it has appropriate loudness - this is the key contribution to this concept by the unimpeded mastering engineer.

Anyway, this is why I'm so confident that this could work... My reference CD is at K-20 and I loudness-adjusted a bunch of tracks including...
The Police - Every Breath You Take (-7)
Bob Marley & The Wailers - Exodus (-8)
Suzanne Vega - Rock In This Pocket (-12)
Sade - King Of Sorrow (-14)
Ricky Martin - She Bangs (-13)

The figure in brackets is the level change in dB needed to bring each one down to work nicely at K-20.

The result is that the more dynamic tracks fit nicely with the slammed tracks.
Exodus is just beautiful in its texture - it's the standard I aim for because when the organ comes in, it feels like you could run your fingers over the sound and feel it.

In contrast, the slammed tracks (particularly She Bangs) are devoid of texture, sound muffled, and lack punch. The slammed tracks are not inappropriately louder than the dynamic ones.

Admittedly I don't have the technology to determine the exact RMS values of these songs but looking at the meters everything seems to be where I expect it. The biggest difference being that the peak line on the slammed tracks doesn't get anywhere as close to 0 dBFS as it does on the dynamic tracks. If anything, I can see the effects of the excessive limiting almost as much as I can hear it.

My conclusion is that not only will dynamic tracks and slammed tracks coexist in the face of a standard that limits RMS levels but it is the slammed tracks (not the dynamic) that will actually be disadvantaged in the ear of the listener. The difference in quality at K-20 is enough that ordinary people would notice (again excluding ourselves and HiFi buffs from being "ordinary"). At K-14 the difference won't be quite so obvious but ordinary people would definitely be able to hear the difference between dynamic and slammed when their attention is drawn to it.

 

[dualflip]

Well, i dont think the loudness wars will be over, or at least anytime soon. I think that music represents the times we live in, so does the sound of it, we are in the "fast-food" era, everything is plastic, everything is disposable, it doesnt matter if its any good as long as is cheap and lasts long enough to replace with a new one within a 6-month to 1 year period. Society is stressed up all the time, the world is running all the time, everything is loud: cities, malls, buildings, etc... So i think the loudness wars have its root based on something else than just "Radio sound", its the reflection of our society these days, so why not embrace it?, i mean im totally against the loudness wars, but as i said we live in fu**ed up times, loudness is just what will define us as an era for the generations to come, so in 20 years from now if society is different, and dropes down the hysteria (i hope this is over, even sooner if possible) when people hear a record made in this times, they will have a good picture of how the world was those days, just like when you hear 60's and 70's albums and you can imagine the hippie era, vietnam war, etc..

Crushing things, loudness, etc.. is something very nasty and fu***d up, but if you think of it... we also are...

 

[SouthSIDE Glen]

Now you're confusing me. Our stuff with RMS of -14 and crest factor of 14 will sound approximately as loud as stuff with RMS of -14 and crest factor of 5.

I guess a lot here depends upon the agreed upon definitions of "RMS" and"Peak RMS". One of the problems with some RMS measurements is that the size of the time slice over which the average is measured can vary.

This is what I was thinking when I read the terminology for the Katz proposal that talked about "peak RMS" being used as the gauge. This - correct me if I'm wrong - means that -14 would not be the overall RMS, or overall average volume, of the track, but rather the peak RMS level as read by the individual ongoing RMS measurements as the song is sampled.

If this is the case, then it is entirely possible that a highly dynamic track can have a peak RMS of - 14 yet the overall perceived volume of the track would be lower than a track which has smashed it's dynamics and has a fairly consistent RMS level of -14.

And throwing a slightly different focus on the issue, because of the very nature of RMS measurement, if you have two tracks of entirely different sonic density - a sparse ballad vs. a wall of sound anthem - there is no guarantee that balancing volume by perceived volume is going to wind up with matching RMS numbers. Put another way, an RMS measurement is not automatically a measurement of perceived volume. Often times, the ballad needs to ride several dB below the anthem - at least according to the numbers - in order to sound "matched" in volume.

You're right, tonality can make a difference, which is important to this discussion. Even more so, the sonic density of the mix can effect it even more.

I'd be curious; on those tracks you have listed, how well do your volume change numbers actually match the measured differences in both peak and overall RMS values? I could be wrong, but I'd be fairly surprised if there was a consistent 1:1 correlation between original RMS readings (both peak and overall) and the amount of gain change you need to apply to match them in perceived playback volume.

G.

 

[iqi616]

Ah, I'm with you now Glen. So the quiet sections of the slammed tracks would be too loud in comparison to the quiet sections of the dynamic tracks (being that the upper limit is pegged). Yes, that's a problem that I don't really have an answer to. I think it applies whether we're talking RMS or peak RMS.

Having accepted this as being one of several problems, I don't think it invalidates the aim of having a standard. The noticeable variation between slammed and dynamic tracks would be far less than it is right now. The listener wouldn't be having the aversion therapy they get now when a slammed track follows a dynamic track on their iPod - it's amazing how quick you can yank the earbuds out of your ears when you need to!

The other problems I can see is that a lot of recording folks will get hung up on the numbers and look to technology to make the decisions for them when really it is down to the mastering engineers making decisions on the loudness of each track like they did in the old days. The loudest track was pegged by the technology but the rest were levelled by ear.

Or worse still, it'll all get bogged down with endless discussions over whether the limit should be 14 dB or 13 dB or 15 dB and nothing will happen. That's my greatest fear because even a poorly implemented standard would be better than the current de-facto standard where the upper limit is 0 dBFS.

 

[iqi616]

I'd be curious; on those tracks you have listed, how well do your volume change numbers actually match the measured differences in both peak and overall RMS values? I could be wrong, but I'd be fairly surprised if there was a consistent 1:1 correlation between original RMS readings (both peak and overall) and the amount of gain change you need to apply to match them in perceived playback volume.

Unfortunately I use multitrackers rather than a PC so I don't have that technology. Is there any economically priced (i.e. cheap ) PC software out there that I could check this out with? Any plug-ins I could use with Audacity?

 

[suprstar]

*Sigh*... no there *aren't* really any valid reasons, it's almost all downside.

Whoa, calm down, I'm on your side! But yes, there are, if there wasn't, nobody would do it. Relatively loud songs/commercials catch your attention, that's a fact. If you're in marketing/promotion/whatever, that's better than a masterpiece that goes unnoticed. It's also fact that you pay for that loudness with sound quality. We all know SQ is obviously not top priority for a lot of decision makers.. As a musician and (wannabe) producer, I'm certainly no advocate of the war,

 

[SouthSIDE Glen]

Whoa, calm down, I'm on your side! But yes, there are, if there wasn't, nobody would do it.

People act as a collective based upon unsound reasoning all the time; myths take on a life of their own. If you doubt that, take one look at Washington and Wall Street right now. It turns out that common sense people like you and I who have been saying for twenty years now that there's no way that all these people could actually afford all these McMansions were right. And now our economy is in big trouble because people bought into the illusion that common sense didn't matter.

Before that it was the idiotic Internet bubble. Remember the Super Bowl when every $3M commercial that wasn't for Budweiser was for some website that wasn't even worth $3M and wasn't even making any money? The world is full of examples where everybody buys into a mass illusion that is just plain wrong from the get go.

Remember where the term "Don't drink the Koolaid" came from...

Relatively loud songs/commercials catch your attention, that's a fact.

The key word there is "relatively". When everyone is loud, no one is loud. When everyone is loud, the only thing that catches your attention is the thing that *isn't* as loud as the rest.

Besides, catching your attention is worthless if you don't like the music. And if the music is good, it doesn't have to be loud to get your attention.

And we haven't even started talking abut the fatigue factor; that the more crushed the music, the faster it tends to drop off of people's playlists because of actual listening fatigue. BTW, you think this fact is lost on the labels, A&R guys and the producers whose income is based on cranking as much new material through the grinder as possible? Yeah, they want to grab your attention alright; they want to grab your attention this week with this crap production to replace last week's crap production.

The problem is that a) you, as the musician, can never make it over the contract hump that way, and 2) the music industry faces the same danger of implosion that the subprimes, the dotcoms and the Jonesies eventually experienced.

G.

 

[SouthSIDE Glen]

Unfortunately I use multitrackers rather than a PC so I don't have that technology. Is there any economically priced (i.e. cheap ) PC software out there that I could check this out with? Any plug-ins I could use with Audacity?

Is free cheap enough for you, Mike? For a running RMS meter (amongst other metrics), try the free version of Inspector from Roger Nuchols Digital. I also like the Span plug from Voxengo; it's RMS numbers jive quite well with the numbers I get from within Sound Forge, and both seem to be trustworthy values.

G.

 

[jndietz]

I heard that someone "mystery sound engineer" managed to remix one of the tracks on Death Magnetic off of the Guitar Hero/Rock Band game to make it sound better.

Can anyone confirm this??

 

[iqi616]

Is free cheap enough for you, Mike? For a running RMS meter (amongst other metrics), try the free version of Inspector from Roger Nuchols Digital. I also like the Span plug from Voxengo; it's RMS numbers jive quite well with the numbers I get from within Sound Forge, and both seem to be trustworthy values.

G.

Thanks Glen I'll check those out.

 

[jndietz]

So guys... I have been running my master bus's preamp about +15dB into a MBC and a brickwall limiter... am I doing it right???

 

[Steve Henningsgard]

So guys... I have been running my master bus's preamp about +15dB into a MBC and a brickwall limiter... am I doing it right???

No no no, you've got it all wrong. Just bus everything to a stereo aux channel with a couple of limiters pegged to 100, and then send THAT to the master bus at +15dB. Be sure to put another 3 or 4 limiters, preferably different types to really get rid of every possible dynamic.

You'll be gold.